Three-dimensional phase unwrapping method based on nonlinear phase coding

Abstract

The invention discloses a three-dimensional phase unwrapping method based on nonlinear phase coding, which comprises the following steps of: 1, constructing a wavelength-De Bruijn element mapping table, and generating a non-equal period phase shift stripe group; 2, projecting the phase shift fringes to a measured object frame by frame, obtaining a modulation pattern, and decoding the modulation pattern to obtain a non-linear wrapped phase; 3, constructing an approximate phase to carry out local phase unwrapping on the nonlinear wrapped phase to obtain a nonlinear local phase; and 4, screening out a phase candidate point and a candidate phase order of a nonlinear local phase by using a three-dimensional phase matching method. And step 5, using the phase candidate points to guide the nonlinear local phase to perform phase correlation matching so as to screen out a unique phase order, thereby realizing phase unwrapping. According to the method, a novel phase distortion characteristic is introduced, and a plurality of sine fringes with different wavelengths are used for serial coding according to a DB sequence. In a decoding stage, robust phase disambiguation is achieved by minimizing a local phase difference in left and right views. And high-precision measurement can be realized only by using three projection patterns, so that the three-dimensional measurement efficiency is greatly improved.

Description

Technical Field

[0001] The invention relates to the technical field of three-dimensional shape measurement, and in particular to a three-dimensional phase unwrapping method based on nonlinear phase encoding. Background Art

[0002] With the advancement of image processing, optics and computer technology, optical image processing technology based on optoelectronic imaging theory has developed rapidly. This technology has undergone extensive research and application in the past few decades, and has also promoted major breakthroughs in computer vision technology. Optical image processing technology, with its advantages in information processing and visual perception, has led the continuous evolution of the field of computer vision. With the increasing demand for information richness and diversity, the evolution of computer vision technology has gradually expanded from traditional two-dimensional images to more complex three-dimensional spaces. This transformation requires not only more intelligent and efficient technology, but also the ability to efficiently obtain high-quality spatial stereoscopic visual perception information, which has become one of the current research focuses in the field of computer vision.

[0003] Optical 3D measurement technology has been widely used in the industrial field as a low-cost, high-precision and efficient non-contact measurement method. Robot navigation, industrial modeling, microscopic measurement and other fields rely on the accuracy and reliability of optical 3D measurement technology. The mainstream structured light measurement method includes fringe projection profile measurement, which uses a projector to project a grating image carrying phase information, and then uses a camera to synchronously shoot and phase decode to obtain the surface shape information of the object being measured. However, due to the particularity of phase information, that is, the phase circulates in the range of (-π,π], which leads to phase ambiguity. In order to eliminate this ambiguity, phase unwrapping technology is needed to restore the continuous absolute phase.

[0004] After decades of development, researchers have proposed many effective phase unwrapping methods, including spatial phase unwrapping and temporal phase unwrapping. Spatial phase unwrapping usually uses the phase relationship between pixels to achieve phase unwrapping. However, when there are depth mutations or isolated objects in the scene, phase unwrapping errors often occur. In contrast, the temporal phase unwrapping method uniquely identifies each phase order of the wrapped phase through additional multi-frame data acquisition, thereby solving the limitations of the spatial phase unwrapping method. However, the temporal phase unwrapping method also has some problems, such as limited measurement speed and accumulation of time domain errors. Therefore, one of the current research focuses is to develop a reliable phase unwrapping method that does not require additional modes to meet the requirements of accuracy and efficiency in practical applications. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide a three-dimensional phase unwrapping method based on nonlinear phase encoding.

[0006] The technical solution of the invention is as follows:

[0007] A three-dimensional phase unwrapping method based on nonlinear phase encoding comprises the following steps:

[0008] Step 1: Construct a wavelength-De Bruijn element mapping table to generate a non-uniform periodic phase-shifted fringe group;

[0009] Step 2: Project the fringe images in the phase-shifted fringe group onto the object to be measured frame by frame, obtain the modulation pattern and decode it to obtain the nonlinear wrapped phase;

[0010] Step 3: Construct an approximate phase to perform local phase unwrapping on the nonlinear wrapped phase to obtain a nonlinear local phase;

[0011] Step 4: Use the stereo phase matching method to screen out candidate phase points and candidate phase orders of the nonlinear local phase;

[0012] Step 5: Use the phase candidate points to guide the nonlinear local phase to perform phase correlation matching to screen out the unique phase order, thereby achieving phase unwrapping.

[0013] Furthermore, the non-uniform periodic phase-shifted fringe group in step 1 is obtained by the following steps:

[0014] Step 1.1, select m wavelengths and De Bruijn elements, and construct a wavelength-De Bruijn element mapping table;

[0015] Step 1.2, generate an m-element n-order De Bruijn wavelength sequence and generate non-uniform periodic phase shift coding; m∈Ν + , n∈Ν + , N + represents a set of positive integers;

[0016] The wavelength-De Bruijn element mapping table in step 1.1 is:

[0017] De Bruijn Element <![CDATA[v0]]> <![CDATA[v1]]> ... <![CDATA[v m-1 ]]> wavelength <![CDATA[f0]]> <![CDATA[f1]]> ... <![CDATA[f m-1 ]]>

[0018] The De Bruijn wavelength sequence in step 1.2 is:

[0019]

[0020] Where m∈Ν + , n∈Ν + , N + Represents the set of positive integers.

[0021] The non-equal period phase shift encoding formula in step 1.2 is:

[0022]

[0023] in

[0024]

[0025] u and v represent the horizontal and vertical coordinates of the projector pixel plane, 1≤u≤U, 1≤v≤V; U and V represent the horizontal and vertical pixel resolutions of the projector DMD surface, respectively. is a non-uniform periodic phase shift pattern, N represents the number of phase shift patterns, N≥3, i∈{1,2,...,N}. It can be considered as a function of the pixel coordinate v, i.e. the first row of pixels in the coded pattern. There is no u on the right side, which means that the pixel distribution of each row from u=1 to u=U is the same (1≤u≤U). Represents an image with U rows and V columns.

[0026] In step 1.2 The elements in . is not represented individually; it is Sum operation The elements in F DB .

[0027] The nonlinear wrapped phase in step 2 is obtained as follows:

[0028] The projector projects a non-uniform periodic phase-shifted fringe pattern onto the scene to be tested, and the industrial camera synchronously captures the deformed fringe pattern, which is then further calculated to obtain the nonlinear wrapping phase.

[0029] The deformed fringe pattern in step 2 is represented as:

[0030]

[0031] Where x and y represent the horizontal and vertical coordinates of the camera pixel plane, 1≤x≤W, 1≤y≤H. W and H represent the horizontal and vertical pixel resolutions of the camera CCD surface, respectively. α(x,y) represents the reflectivity of the object; β1(x,y) represents the intensity of the ambient light on the surface of the object, and β2(x,y) represents the intensity of the ambient light directly entering the camera from the environment. Indicates the distorted fringe pattern captured by the camera. DB (x,y) is the average pixel intensity of the entire pattern set, B DB (x,y) represents the intensity modulation of a given pixel; φ DB (x,y) represents the nonlinear wrapping phase; i = 1,...,N;

[0032] The nonlinear wrapping phase calculation formula in step 2 is:

[0033]

[0034] The nonlinear local phase in step 3 is obtained as follows:

[0035] Step 3.1, calculating the nonlinear wrapping phase gradient and the nonlinear absolute phase gradient of the nonlinear wrapping phase along the x increasing direction;

[0036] Step 3.2, construct an approximate phase using the nonlinear phase gradient, and use the approximate phase to unwrap the nonlinear local phase to obtain the nonlinear local phase;

[0037] The nonlinear wrapping phase gradient calculation formula in step 3.1 is:

[0038]

[0039] Among them, ▽ represents the gradient operator, represents the nonlinear wrapped phase gradient along the increasing direction of x, and o represents the pixel product operator. ;

[0040] The nonlinear absolute phase gradient calculation formula in step 3.1 is:

[0041]

[0042] in, represents the nonlinear absolute phase gradient along the increasing direction of x;

[0043] The approximate phase calculation formula in step 3.2 is:

[0044]

[0045] Indicates approximate phase;

[0046] The nonlinear local phase calculation formula in step 3.2 is:

[0047]

[0048] in

[0049]

[0050] M>(n-1)×max{f0,f1,...f m-1}, round{.} represents the rounding operator, mod[.] represents the remainder operator, represents the nonlinear local phase, 0≤τ≤M-1, τ∈Ν + .

[0051] The phase candidate points in step 4 are obtained as follows:

[0052] Step 4.1, after the industrial cameras on both sides of the projector acquire the image, step 2 is performed respectively to obtain the wrapping phase in the left and right image planes;

[0053] Step 4.2, estimating the potential absolute phase of the left image plane wrapped phase, and calculating the encoded horizontal coordinate of the potential absolute phase;

[0054] Step 4.3, construct the three-view mapping equation, calculate the wrapped phase candidate points of the right image plane, perform left and right phase consistency detection to screen out the phase candidate point set and candidate phase order set;

[0055] The wrapping phases in the left and right image planes calculated in step 4.1 are expressed as and Where (x L ,y L ) and (x R ,y R ) represent the horizontal and vertical coordinates of the left and right image planes respectively;

[0056] The calculation formula for the potential absolute phase of the left image plane wrapped phase in step 4.2 is:

[0057]

[0058] Among them, k(x L ,y L ) represents the phase order of the left view, k∈{0,1,...,m n -1}, Indicates that (x L ,y L ) coordinate position potential k-th absolute phase value. Next, calculate The potential encoding abscissa in the corresponding projector view is calculated as:

[0059]

[0060] in

[0061]

[0062] floor(.) is the floor operator, mod(.) is the remainder operator, u k It represents the potential horizontal coordinate of the pixel point in the left image plane mapped to the corresponding point under the projector image plane, and H[.] is the piecewise distortion correction operator;

[0063] The three-view mapping equation in step 4.3 is:

[0064]

[0065] The t operator represents the transposition calculation of the matrix, and s L 、s R and P represents the scaling factor of the left and right cameras and the projector, (u k ,v k ) represents the phase point in the left image plane The potential horizontal and vertical coordinates of the corresponding points mapped to the projector image plane. L , P R and P P Represent the projection matrices of the left and right cameras and the projector respectively, which are obtained using camera calibration technology (see reference: Zhang ZA Flexible New Technique for Camera Calibration[J].IEEE Transactions on Pattern Analysis and Machine Intelligence, 2000, 22(11):1330-1334.DOI:10.1109 / 34.888718.). express The corresponding wrapped phase candidate points of the right image plane. Therefore, the three-view mapping equation is further simplified to

[0066]

[0067] Among them, the function f lr The function of [.] is to map the absolute phase in the left view to the coordinates in the right view. After combining all the equations in steps 4.2-4.3, we can achieve arrive Calculation of f lr [.] represents the abbreviated joint process;

[0068] The calculation formula for the wrapped phase candidate point set and candidate phase order set of the right image plane in step 4.3 is:

[0069]

[0070] in

[0071]

[0072] Ω p represents the set of selected phase candidate points, Ω k represents the set of candidate phase orders screened out, Represents the set Ω pThe candidate points in , Δδ is the noise tolerance threshold, Indicates phase deviation.

[0073] The absolute phase in step 5 is obtained as follows:

[0074] Step 5.1, the nonlinear local phase of the left view is combined with different candidate phase orders and mapped to the right view to obtain the candidate local phase of the right view, and the unique phase order is screened out by minimizing the left and right local phase residuals;

[0075] Step 5.2, the phase order is combined with the nonlinear wrapped phase to perform phase unwrapping to obtain the nonlinear absolute phase, and then the linear absolute phase is calculated through piecewise distortion correction;

[0076] The candidate local phase calculation formula in step 5.1 is:

[0077]

[0078] in, Represents each candidate point p in the right view k ∈Ω p The corresponding local phase of represents the nonlinear local phase of the left view; then, combined with and The unique phase order is selected by minimizing the left and right local phase residuals

[0079]

[0080] in

[0081]

[0082] J sum represents the local phase residual, |.| represents the absolute value calculation;

[0083] The nonlinear absolute phase calculation formula in step 5.2 is:

[0084]

[0085] in, Represents the nonlinear absolute phase of the left view.

[0086] The linear absolute phase calculation formula in step 5.2 is:

[0087] Right Wrap Phase Only used for auxiliary left wrap phase The phase unwrapping of the right view does not require the calculation of the absolute phase of the right view. Only a single absolute phase Φ is required for 3D imaging L (x L,y L )

[0088] Among them, Φ L (x L ,y L ) represents the linear absolute phase of the left view with constant gradient, f p Represents the wavelength of the sinusoidal fringes used during the projector calibration phase.

[0089] Beneficial effects:

[0090] The beneficial effects of the present invention are:

[0091] 1. The present invention proposes a three-dimensional phase unwrapping method based on nonlinear phase coding, which uses multiple single-frequency fringes of different frequencies to perform combined coding in the De Bruijn order. The coding is composed of pure sinusoidal fringes and is robust to the complex structure and different reflection areas of the object. Since the amplitude and intensity distribution of the sinusoidal are not modified, high measurement accuracy is guaranteed;

[0092] 2. The present invention uses stereo matching of local phases and minimization of matching errors, and utilizes the uniqueness of phase distortion to achieve fast phase unwrapping without the need for additional patterns and equipment. Compared with image feature matching, local phase matching is more robust and is not affected by image deformation caused by different viewing angles.

[0093] 3. The present invention achieves measurement accuracy similar to that of the advanced multi-frequency heterodyne method using only 1 / 3 of the number of projections, greatly improving the coding efficiency;

[0094] 4. The present invention does not require the introduction of additional equipment and does not require the projection of additional stripes, thereby ensuring low cost and high measurement efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0095] Figure 1 It is a schematic diagram of a phase unwrapping process of a three-dimensional phase unwrapping method based on nonlinear phase encoding in Example 1 of the present invention;

[0096] Figure 2 It is a three-step non-equal period phase shift diagram in Example 1 of the present invention;

[0097] Figure 3 These are three deformed fringe images obtained in Example 1 of the present invention;

[0098] Figure 4 is the nonlinear wrapped phase diagram in Embodiment 1 of the present invention;

[0099] Figure 5 is the nonlinear wrapped phase gradient image in Example 1 of the present invention;

[0100] Figure 6is the nonlinear absolute phase gradient map in Example 1 of the present invention;

[0101] Figure 7 This is an approximate phase diagram in Example 1 of the present invention.

[0102] Figure 8 This is a phase order diagram in Example 1 of the present invention.

[0103] Fig. 9 This is the linear absolute phase diagram in Example 1 of the present invention.

[0104] Fig.10 This is the object model in Example 1 of the present invention. DETAILED DESCRIPTION

[0105] The present invention will be further described below in conjunction with the accompanying drawings and embodiments. However, this should not be understood as the scope of the above subject matter of the present invention being limited to the following embodiments, and all technologies realized based on the content of the present invention belong to the scope of the present invention.

[0106] In order to better illustrate the method of the present invention, this embodiment takes a 5-element 3rd-order De Bruijn sequence as an example, combines a three-step phase shift as an encoding scheme, and illustrates the measurement of a complex plaster figure and a calibration plate.

[0107] like Figure 1 As shown in the flowchart of the present invention, a three-dimensional phase unwrapping method based on nonlinear phase encoding includes the following steps:

[0108] Step 1: Construct a wavelength De Bruijn element mapping table to generate a non-uniform periodic phase-shifted fringe group.

[0109] In order to ensure the uniformity of the calculated phase, five approximate wavelengths are selected for sinusoidal encoding, including: 16, 18, 20, 22 and 24. Then the wavelengths are assigned to five De Bruijn elements to construct a wavelength-De Bruijn element mapping table.

[0110] De Bruijn element-wavelength mapping table

[0111] De Bruijn Element 1 2 3 4 5 wavelength 16 18 20 22 24

[0112] Then, a set of 5-element 3rd-order De Bruijn wavelength sequences is constructed, as shown in formula (1);

[0113]

[0114] Among them, F DB represents a 5-ary 3rd order De Bruijn sequence. Next, use F DBThe wavelength in generates a three-step De Bruijn phase shift fringe pattern, and the calculation formula is:

[0115]

[0116] in

[0117]

[0118] (u,v) represents the horizontal and vertical coordinates of the projector plane, i∈{1,2,3}. , is a non-uniform periodic phase shift pattern, such as Figure 2 shown.

[0119] Step 2: Project the phase-shifted fringes frame by frame, obtain the modulation pattern and decode it to get the nonlinear wrapped phase.

[0120] The projector projects a three-step De Bruijn phase-shifted fringe pattern onto the scene to be tested, and the industrial camera simultaneously captures the deformed fringe pattern. like Figure 3 As shown. The nonlinear wrapping phase is further calculated using formula (4), as Figure 4 As shown;

[0121]

[0122] Step 3: Construct an approximate phase and locally expand the nonlinear wrapped phase to obtain the nonlinear local phase.

[0123] First, the nonlinear wrapping phase φ is calculated using formula (5) DB (x,y) Gradient along the increasing x direction.

[0124]

[0125] Among them, ▽ represents the gradient operator, represents the nonlinear wrapped phase gradient along the increasing y direction, such as Figure 5 As shown in Figure 6, the phase gradient of the nonlinear wrapped phase has a phase truncation of 2π, which causes periodic jumps in the phase gradient. Therefore, we use formula (6) to Do gradient correction

[0126]

[0127] in, represents the nonlinear absolute phase gradient along the increasing direction of x, such as Figure 6 Next, use formula (7) to calculate the approximate absolute phase

[0128]

[0129] Represents the approximate phase, such as Figure 7 Next, use To assist DB (x, y) to perform local phase unwrapping and use formula (8) to calculate φ DB Local phase unwrapping is performed on a pixel region of size 1×M in (x,y).

[0130]

[0131] in

[0132]

[0133] M>(n-1)×max{f0,f1,...f m-1}, round{.} represents the rounding operator, mod[.] represents the remainder operator, represents the nonlinear local phase.

[0134] Step 4: Use the stereo phase matching method to screen out phase candidate points.

[0135] After the industrial cameras on both sides of the projector acquire the image, step 2 is performed respectively, and the calculated wrapping phases in the left and right image planes are expressed as and Where (x L ,y L ) and (x R ,y R ) represent the horizontal and vertical coordinates of the left and right image planes respectively. L ,y L )middle, With 5 3 = 125 possible phase orders, whose corresponding potential absolute phases are calculated using formula (11).

[0136]

[0137] Among them, k(x L ,y L ) represents the left view phase order, k∈{0,1,...,124}, Indicates that (x L ,y L ) coordinate position potential k-th absolute phase value. Next, calculate The corresponding potential encoding abscissa is calculated as:

[0138]

[0139] in

[0140]

[0141] floor(.) is the floor operator, u k It represents the potential horizontal coordinate of the pixel point in the left image plane mapped to the corresponding point under the projector image plane. Next, take out the pre-calibrated parameter P L , P R and P P , construct the three-view mapping equation of the system:

[0142]

[0143] The t operator represents the transposition calculation of the matrix, and s L 、s R and P represents the scaling factor of the left and right cameras and the projector, (u k ,v k ) represents the phase point in the left image plane The potential horizontal and vertical coordinates of the corresponding points mapped to the projector image plane. L , P R and P P Represent the projection matrices of the left and right cameras and the projector respectively, which are obtained using camera calibration technology (see reference: Zhang Z. A Flexible New Technique for Camera Calibration [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2000, 22 (11): 1330-1334. DOI: 10.1109 / 34.888718.). express The corresponding wrapped phase candidate point of the right image plane. Therefore, the combined formulas (12-16) can construct the following mapping

[0144]

[0145] Among them, the function f lr The function of [.] is to map the absolute phase in the left view to the coordinates in the right view. Next, the left and right phase consistency detection is performed to select a small number of phase candidate points. The calculation formula is:

[0146]

[0147] in

[0148]

[0149] Ω prepresents the set of selected phase candidate points, Ω k represents the set of candidate phase orders screened out, Represents the set Ω p The candidate points in , Δδ is the noise tolerance threshold, Indicates phase deviation.

[0150] Step 5: Local phase correlation matching is used to select a unique phase order and achieve phase unwrapping.

[0151] Since the encoding in this patent has nonlinear characteristics, each candidate point p in the right view k ∈Ω p The local phase surface of the left view has different nonlinear distortions. Therefore, only the correct phase order can make the local phase of the left view and right view local wrapping phase The matching consistency is satisfied. The calculation formula is

[0152]

[0153] in represents the nonlinear local phase of the left view. Next, combine and A unique phase order is obtained by minimizing the matching error of the local phases of the left and right views.

[0154]

[0155] in

[0156]

[0157] k(x L ,y L )like Figure 8 Next, the nonlinear wrapping phase φ DB (x L ,y L ) combined with k(x L ,y L )Use formula (23) for phase unwrapping.

[0158] Φ DB (x L ,y L )=φ DB (x L ,y L )+2π×k(x L ,y L ) (twenty three)

[0159] Among them, Φ DB (xL ,y L ) represents the nonlinear absolute phase of the left view. Next, the nonlinear absolute phase is mapped into a linear absolute phase.

[0160]

[0161] Among them, Φ L (x L ,y L ) represents the linear absolute phase of the left view with a constant gradient, such as Fig. 9 As shown. p represents the wavelength of the sinusoidal fringes used in the projector calibration stage. Finally, the linear absolute phase is combined with the structured light system parameters to reconstruct the object model (see reference: Li, Beiwen, et al. "High-accuracy, high-speed 3D structured light imaging techniques and potential applications to intelligent robotics." International journal of intelligent robotics and applications 1.1 (2017): 86-103.), such as Fig.10 shown.

Claims

1. A three-dimensional phase unwrapping method based on nonlinear phase encoding, characterized in that: The following steps are involved: Step 1: Construct a wavelength-De Bruijn element mapping table to generate a non-uniform periodic phase-shifted fringe group; Step 2: Project the fringe images in the phase-shifted fringe group onto the object to be measured frame by frame, obtain the modulation pattern and decode it to obtain the nonlinear wrapped phase; Step 3: Construct an approximate phase to perform local phase unwrapping on the nonlinear wrapped phase to obtain a nonlinear local phase; Step 4: Use the stereo phase matching method to screen out candidate phase points and candidate phase orders of the nonlinear local phase; Step 5: Use the phase candidate points to guide the nonlinear local phase to perform phase correlation matching to screen out the unique phase order, thereby achieving phase unwrapping.

2. The three-dimensional phase unwrapping method based on nonlinear phase encoding according to claim 1, characterized in that: The non-uniform periodic phase-shifted fringe group in step 1 is obtained by the following steps: Step 1.1, select m wavelengths and De Bruijn elements, and construct a wavelength-De Bruijn element mapping table; Step 1.2, generate an m-element n-order De Bruijn wavelength sequence and generate non-uniform periodic phase shift coding; m∈Ν + , n∈Ν + , N + represents a set of positive integers; The wavelength-De Bruijn element mapping table in step 1.1 is: The De Bruijn wavelength sequence in step 1.2 is: Where m∈Ν + , n∈Ν + , N + Represents the set of positive integers. The non-equal period phase shift encoding formula in step 1.2 is: in u and v represent the horizontal and vertical coordinates of the projector pixel plane, 1≤u≤U, 1≤v≤V; U and V represent the horizontal and vertical pixel resolutions of the projector DMD surface, respectively. is a non-uniform periodic phase shift pattern, N represents the number of phase shift patterns, N≥3, i∈{1,2,...,N}.

3. The three-dimensional phase unwrapping method based on nonlinear phase encoding according to claim 1, characterized in that: The nonlinear wrapped phase in step 2 is obtained as follows: The projector projects a non-uniform periodic phase-shifted fringe pattern onto the scene to be tested, and the industrial camera synchronously captures the deformed fringe pattern, which is then further calculated to obtain the nonlinear wrapping phase. The deformed fringe pattern in step 2 is represented as: Where x and y represent the horizontal and vertical coordinates of the camera pixel plane, 1≤x≤W, 1≤y≤H. W and H represent the horizontal and vertical pixel resolutions of the camera CCD surface, respectively. α(x,y) represents the reflectivity of the object; β1(x,y) represents the intensity of the ambient light on the surface of the object, and β2(x,y) represents the intensity of the ambient light directly entering the camera from the environment. Indicates the distorted fringe pattern captured by the camera. DB (x,y) is the average pixel intensity of the entire pattern set, B DB (x,y) represents the intensity modulation of a given pixel; φ DB (x,y) represents the nonlinear wrapping phase; i = 1,...,N; The nonlinear wrapping phase calculation formula in step 2 is:

4. The three-dimensional phase unwrapping method based on nonlinear phase encoding according to claim 1, characterized in that: The nonlinear local phase in step 3 is obtained as follows: Step 3.1, calculating the nonlinear wrapping phase gradient and the nonlinear absolute phase gradient of the nonlinear wrapping phase along the x increasing direction; Step 3.2, construct an approximate phase using the nonlinear phase gradient, and use the approximate phase to unwrap the nonlinear local phase to obtain the nonlinear local phase; The nonlinear wrapping phase gradient calculation formula in step 3.1 is: in, represents the gradient operator, represents the nonlinear wrapped phase gradient along the increasing direction of x, and o represents the pixel product operator; The nonlinear absolute phase gradient calculation formula in step 3.1 is: in, represents the nonlinear absolute phase gradient along the increasing direction of x; The approximate phase calculation formula in step 3.2 is: Indicates approximate phase; The nonlinear local phase calculation formula in step 3.2 is: in M>(n-1)×max{f0,f1,...f m-1 }, round{.} represents the rounding operator, mod[.] represents the remainder operator, represents the nonlinear local phase, 0≤τ≤M-1, τ∈Ν + .

5. The three-dimensional phase unwrapping method based on nonlinear phase encoding according to claim 1, characterized in that: The phase candidate points in step 4 are obtained as follows: Step 4.1, after the industrial cameras on both sides of the projector acquire the image, step 2 is performed respectively to obtain the wrapping phase in the left and right image planes; Step 4.2, estimating the potential absolute phase of the left image plane wrapped phase, and calculating the encoded horizontal coordinate of the potential absolute phase; Step 4.3, construct the three-view mapping equation, calculate the wrapped phase candidate points of the right image plane, perform left and right phase consistency detection to screen out the phase candidate point set and candidate phase order set; The wrapping phases in the left and right image planes calculated in step 4.1 are expressed as and Where (x L ,y L ) and (x R ,y R ) represent the horizontal and vertical coordinates of the left and right image planes respectively; The calculation formula for the potential absolute phase of the left image plane wrapped phase in step 4.2 is: Among them, k(x L ,y L ) represents the phase order of the left view, k∈{0,1,...,m n -1}, Indicates that (x L ,y L ) coordinate position potential k-th absolute phase value. Next, calculate The potential encoding abscissa in the corresponding projector view is calculated as: in floor(.) is the floor operator, mod(.) is the remainder operator, u k It represents the potential horizontal coordinate of the pixel point in the left image plane mapped to the corresponding point under the projector image plane, and H[.] is the piecewise distortion correction operator; The three-view mapping equation in step 4.3 is: The t operator represents the transposition calculation of the matrix, and s L 、s R and P represents the scaling factor of the left and right cameras and the projector, (u k ,v k ) represents the phase point in the left image plane The potential horizontal and vertical coordinates of the corresponding points mapped to the projector image plane. L , P R and P P Represent the projection matrices of the left and right cameras and the projector, respectively, which are obtained using camera calibration technology; express The corresponding wrapped phase candidate points of the right image plane; simplify the three-view mapping equation to Among them, the function f lr The function of [.] is to map the absolute phase in the left view to the coordinates in the right view; The calculation formula for the wrapped phase candidate point set and candidate phase order set of the right image plane in step 4.3 is: in Ω p represents the set of selected phase candidate points, Ω k represents the set of candidate phase orders screened out, Represents the set Ω p The candidate points in , Δδ is the noise tolerance threshold, Indicates phase deviation.

6. The three-dimensional phase unwrapping method based on nonlinear phase encoding according to claim 1, characterized in that: The absolute phase in step 5 is obtained as follows: Step 5.1, the nonlinear local phase of the left view is combined with different candidate phase orders and mapped to the right view to obtain the candidate local phase of the right view, and the unique phase order is screened out by minimizing the left and right local phase residuals; Step 5.2, the phase order is combined with the nonlinear wrapped phase to perform phase unwrapping to obtain the nonlinear absolute phase, and then the linear absolute phase is calculated through piecewise distortion correction; The candidate local phase calculation formula in step 5.1 is: in, Represents each candidate point p in the right view k ∈Ω p The corresponding local phase is represents the nonlinear local phase of the left view; then, combined with and The unique phase order is selected by minimizing the left and right local phase residuals in J sum represents the local phase residual, |.| represents the absolute value calculation; The nonlinear absolute phase calculation formula in step 5.2 is: in, Represents the nonlinear absolute phase of the left view. The linear absolute phase calculation formula in step 5.2 is: Among them, Φ L (x L ,y L ) represents the linear absolute phase of the left view with constant gradient, f p Represents the wavelength of the sinusoidal fringes used during the projector calibration phase.

Images